OUP user menu

★ editor's choice ★

Molecular imaging using contrast-enhanced ultrasound: evaluation of angiogenesis and cell therapy

Howard Leong-Poi
DOI: http://dx.doi.org/10.1093/cvr/cvp248 190-200 First published online: 23 July 2009


The field of regenerative medicine and its applications for cardiovascular diseases continues to grow rapidly, fuelled by the increasing numbers of symptomatic patients who are not candidates for conventional revascularization procedures and remain refractory to maximal medical therapy. Therapeutic angiogenesis, initially in the form of the administration of growth factor protein or gene therapy and, more recently, in the form of adult progenitor cell therapy, has emerged as a promising new method of treatment for patients with ischaemic heart disease and peripheral arterial disease. There is a growing interest in non-invasive imaging techniques to evaluate the response to angiogenic gene-and cell-based therapies. Contrast-enhanced ultrasound (CEU) techniques using site-specific microbubbles have recently been developed for the molecular imaging of the vascular phenotype that characterizes angiogenesis. These methods have now been modified to allow the imaging of progenitor cell engraftment into neovessels. These molecular imaging techniques using contrast ultrasound and targeted microbubbles have the potential to further characterize the angiogenic response, aid in the optimization of gene- and cell-based strategies of therapeutic neovascularization, and ultimately serve to monitor the therapeutic effects in patients enrolled in clinical trials of regenerative therapies. This review will focus specifically on CEU molecular imaging techniques for the evaluation of angiogenesis and cell therapies in cardiovascular diseases, including: (i) an overview of the techniques and results of pre-clinical studies; (ii) a comparison of CEU molecular imaging techniques with other available molecular imaging modalities; and (iii) a discussion of the future role of CEU molecular imaging in the field of regenerative medicine.

  • This article differs from the Advance Access version: the publication details of reference 46 have been updated to give the issue citation.

  • Ultrasound
  • Molecular imaging
  • Contrast agents
  • Angiogenesis
  • Cell therapy

1. Introduction

Neovascularization occurs by three distinct but potentially overlapping mechanisms: arteriogenesis, the remodelling of newly formed or pre-existing collateral vessels;1 vasculogenesis, the formation of blood vessels from endothelial progenitor cells (EPCs) that traditionally occur during embryonic development2 but has also been noted post-natally;3 and angiogenesis, broadly defined as the growth, development, and maturation of new blood vessels from pre-existing vasculature.4 As such, arteriogenesis, neovasculogenesis, and angiogenesis represent important targets for novel therapeutic strategies. Angiogenesis plays a critical role in the pathophysiology of several disease processes,5 including ocular proliferative diseases, tumour growth, and metastasis and in chronic ischaemic coronary artery disease (CAD) and peripheral arterial disease (PAD).6 Although anti-angiogenic strategies are being developed to suppress tumour growth and prevent metastasis, strategies are being developed to ‘promote’ angiogenesis in the setting of chronic ischaemic heart and peripheral vascular disease.7 Despite ongoing advances in percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) surgical techniques, the number of patients with symptomatic CAD and PAD who are not amenable to revascularization continues to grow. In addition, in many patients undergoing conventional treatment (i.e. PCI or CABG surgery), complete revascularization cannot be achieved, resulting in worse outcomes and angina-free survival rates in CAD patients, and refractory symptoms and limb amputation in PAD patients. In this patient population, novel strategies to promote angiogenesis within ischaemic tissue have the potential to increase blood flow, relieve symptoms, and improve cardiovascular outcomes.

2. Strategies for therapeutic angiogenesis

Although pre-clinical studies and initial small open-labelled trials of angiogenic growth factors and gene therapy have yielded promising results, to date all larger double-blinded randomized placebo-controlled clinical trials have failed to conclusively show a clinical benefit.8,9 Given the lack of clinical success with strategies involving angiogenic protein and gene delivery, the focus has shifted to novel therapies using various adult ‘stem’ or ‘progenitor’ cells, such as EPCs and mesenchymal stem cells (MSCs), to regenerate damaged tissue and promote angiogenesis and neovasculogenesis.10,11 Research in this field of regenerative medicine has progressed very rapidly. Progenitor cell therapy has moved from the bench to the bedside, where early clinical trials have indicated that stem cell therapy may be feasible and beneficial in patients. Although preliminary efficacy data indicate that stem cells have the potential to enhance myocardial perfusion and/or contractile performance in patients with acute myocardial infarction, large outcome trials, particularly in the setting of chronic ischaemic heart disease and heart failure, still need to be conducted.

As the field of regenerative medicine continues to surge forward, several important questions remain unanswered. The promise of gene delivery for therapeutic angiogenesis remains unfulfilled. Although the largely disappointing results of clinical trials of gene therapy for angiogenesis can be explained by several factors, including the choice of the optimal biological agent or combinations of agents and suboptimal delivery techniques, the inability to monitor and assess the efficacy of angiogenic therapies in patients remains a critical limitation.9 A robust technique capable of monitoring the effects of gene therapy on the angiogenic response would play an important role in the optimization of gene delivery strategies for therapeutic neovascularization. For progenitor cell-based strategies, the optimal method of cell delivery, the most beneficial cell type, and the most effective strategy to promote engraftment is currently unknown. Furthermore, the importance of vascular engraftment in the angiogenic response to cell therapy has been questioned. A non-invasive imaging technique capable of spatially and temporally assessing progenitor cell engraftment would help determine the biological fate of injected stem cells into ischaemic tissue, thus providing unique insights into the underlying mechanisms for the functional benefits observed with stem cell therapy.12 Importantly, as clinical trials move forward, the ability to image engrafted stem cells would allow the non-invasive assessment and monitoring of cell-based therapies in patients.13

3. Role of molecular imaging in the evaluation of angiogenesis and cell therapy

Molecular imaging techniques which allow the non-invasive assessment of pathophysiological processes at the molecular or cellular level could be useful to monitor the angiogenic response to gene- and cell-based therapies for neovascularization14 and to track the fate of delivered stem or progenitor cells.12 Although standard imaging tests such as echocardiographic and magnetic resonance imaging (MRI) of left ventricular ejection fraction, MRI spectroscopy of tissue perfusion by measuring the release of ischaemia metabolites such as lactate, and single photon emission computed tomography (SPECT) imaging for regional myocardial perfusion can provide the main functional endpoints of studies of regenerative therapies, information derived from molecular imaging studies could provide equally important and complementary data. The ability to assess the early molecular changes that occur in angiogenesis, prior to resultant changes in perfusion or vascular morphology, would allow early monitoring of therapies. In pre-clinical studies, these techniques could also provide data on the physiological response to angiogenic and cell-based therapies, yielding further insight into mechanisms underlying the pro-angiogenic response. Importantly, these imaging techniques could serve as study endpoints to test novel treatment strategies in the pre-clinical setting and to assess and guide therapies in patients enrolled in clinical trials.

Non-invasive techniques to image angiogenesis and monitor cell engraftment are currently being developed in almost all imaging modalities (Table 1). Studies to date have focused predominantly on MRI15,16 and radionuclide techniques, such as SPECT and positron emission tomography (PET) imaging.17,18 Echocardiography utilizes ultrasound to image cardiac structure and function, and has good spatial and excellent temporal resolution. The introduction of commercially available microbubble contrast agents has now allowed the ultrasonic assessment and quantification of myocardial and tissue perfusion, an important endpoint for studies of therapeutic angiogenesis. The development of dedicated contrast-specific imaging algorithms that suppress background tissue signal and maximize signal from microbubbles has resulted in a higher ‘signal-to-noise’ ratio and a greater sensitivity for the detection of microbubbles.19

View this table:
Table 1

Comparison of techniques for the molecular imaging of angiogenesis and cell therapy

ModalityContrast agentAdvantagesDisadvantages
Ultrasound, echocardiographyMicrobubbles (2–5 µm), with various shell (e.g. lipid, polymer) and gas (e.g. nitrogen, octafluoropropane) compositionWidely available, no ionizing radiation, rapid acquisition, good sensitivity, high temporal resolutionLimited anatomic access, quantification can be difficult, predominantly limited to intravascular targets, imaging stem cells may require genetic modification—potential risk of mutagenesis/immunogenicity
Single photon emission computed tomography (SPECT)High-energy γ emitters (e.g. 123I, 111In, 99mTc)High sensitivity, potential for 3D scanning, wide anatomic access, widely availableIonizing radiation, imaging stem cells may require genetic modification—potential risk of mutagenesis/immunogenicity, modest spatial resolution, quantification can be difficult
Positron emission tomography (PET)High-energy positron emitters (e.g. 18F, 124I, 64Cu)High sensitivity, potential for 3D scanning, wide anatomic access, quantification possibleIonizing radiation, imaging stem cells may require genetic modification—potential risk of mutagenesis and immunogenicity, modest spatial resolution, less widely available
Computed tomography (CT)High-atomic number materials (e.g. iodine)Widely available, 3D scanning, wide anatomic access, wide anatomic access, high spatial resolutionIonizing radiation, potential for nephrotoxicity, artefacts from cardiac devices
Magnetic resonance imaging (MRI)Lanthanides (e.g. gadolinium), superparamagnetic iron-oxide nanoparticles3D scanning, wide anatomic access, high spatial resolution, no ionizing radiation, quantification possibleContraindicated with many intracardiac devices, concern of nephrotoxicity, relatively low sensitivity for cell tracking—related in part to tracer uptake by non-progenitor cells (i.e. macrophages) after cell death, concerns over effects of labelling on cell proliferation and differentiation
Optical imaging: fluorescence and bioluminescenceNear infrared fluorophores; luciferase substratesHigh sensitivity, no ionizing radiation, rapid acquisitionLimited to small animal use due to low depth penetration, imaging stem cells requires genetic modification—risk of mutagenesis/immunogenicity, modest spatial resolution, less widely available, uptake of fluorophore by non-progenitor cells after cell death

Contrast-enhanced ultrasound (CEU) imaging relies on the ultrasonic detection of microbubble contrast agents during their microvascular transit. Molecular imaging using CEU has recently become possible with the development of novel ‘site-targeted’ microbubbles.20,21 Unlike CEU methods to assess tissue perfusion which use free-flowing non-targeted microbubbles, this method uses custom-designed microbubble contrast agents that are retained in regions of disease by virtue of their shell composition/properties or by the conjugation of specific targeting ligands to their outer surface. Since microbubbles remain entirely within the intravascular space throughout their circulation, the processes that can be targeted are characterized by processes that occur primarily within the vascular compartment such as atherosclerosis, ischaemia, and inflammation and are the subject of other invited reviews in this Review Focus on Molecular Imaging in the Cardiovascular System in Cardiovascular Research. After an intravenous injection, site-targeted microbubbles circulate through the vasculature where they have an opportunity to adhere to their specific targets at sites where the processes of interest occur. CEU imaging algorithms have been developed that allow subtraction of tissue signal and signal from any remaining circulating unbound microbubbles, to yield the signal from retained microbubbles alone.22,23 Therefore, CEU molecular imaging is ‘ideally’ suited for the site-targeted imaging of intravascular events, such as neovessel formation during angiogenesis, and progenitor cell incorporation or engraftment in the formation of new vessels.

4. CEU molecular imaging of angiogenesis

Angiogenesis is a complex, multistep process regulated by the interplay of several pro- and anti-angiogenic factors.24 Since microbubble contrast agents are purely intravascular, the specific target for molecular imaging of angiogenesis must be present within the vascular space, and ideally expressed on the luminal surface of the angiogenic vessel. There are several molecules expressed on the endothelial surface of the vasculature that participate directly in the angiogenic process, and can serve as potential endothelial molecular targets for the imaging of angiogenesis (Table 2). Although numerous studies have investigated CEU molecular imaging of angiogenesis in tumour models, fewer experimental studies have examined the potential of CEU molecular imaging of angiogenesis and neovasculogenesis in models of cardiovascular diseases (Table 3). As can be seen, of the potential targets, the majority of experience to date with CEU molecular imaging of angiogenesis has been using microbubbles targeted to αV-integrins. These integrins, specifically αVβ3 and αVβ5, are expressed during angiogenesis and are thought to play a pivotal role in regulating the angiogenic process.25 Microbubble agents targeted to αV-integrins have been developed by conjugation of either monoclonal antibodies (mAbs) targeted against αV-integrins or arginine–glycine–aspartate (RGD) containing peptides that have strong affinity for select integrins, such as αVβ3. Similar mAb and RGD peptide-based strategies to target αV-integrins have allowed targeted autoradiographic, PET, and MRI of tumour vasculature.26,27 Although mAb strategies are less readily translatable to the clinical setting due to issues surrounding immunogenicity, peptide-based targeting strategies suffer from a lack of targeting specificity. For example, although the disintegrin echistatin derived from snake viper venom has strong affinity to integrin αVβ3, it also has affinity to glycoprotein α2bβ3a (IIbIIIa) on activated platelets.28

View this table:
Table 2

Potential molecular targets for CEU molecular imaging of angiogenesis

TargetSpecific moleculeExamples of potential ligandsClinical feasibility and utility/potential issues
Growth factor receptorsReceptors for VEGF, FGF, IGF, TGF-β, Tie-2VEGF121, VEGF165, Ab to VEGF receptorMay provide physiological information on hypoxic stress within tissue. Low targeting efficiency, likely reflecting relatively low absolute level of antigen expression in angiogenic tissues. Concern over immunogenicity of non-humanized antibodies, and slow clearance of monoclonal Ab
IntegrinsαV-integrins (αVβ3—vitronectin receptor)Monoclonal Ab (LM609), RGD containing peptides (e.g. echistatin, bistatin)Highly specific, due to low expression in normal vasculature. Integrins involved in the angiogenic process—theoretical potential for inhibition depending on targeting strategy. May provide information on metastatic potential in tumor imaging. Concern over immunogenicity of non-humanized antibodies, and slow clearance of monoclonal Ab. Peptide-based strategy provides more rapid clearance.
α5β1 (fibronectin receptor)Monoclonal Ab (L-19)
Markers of activated endothelial cells or neovesselsTumour angiogenesis markersAsparagine–glycine–arginine (NGR), arginine–glycine–aspartate (RGD), arginine–arginine–leucine (RRL)Concern over immunogenicity of non-humanized antibodies, and slow clearance of monoclonal Ab. Peptide-based strategy provides more rapid clearance. May provide physiological information on underlying mechanisms of angiogenesis
CD105 (endoglin)Ab to CD105
VCAM-1Ab to VCAM-1
Activated endothelial cellsAngiostatin, endostatin
View this table:
Table 3

Experimental studies of CEU molecular imaging of angiogenesis and neovasculogenesis in vascular models

StudyMicrobubble composition/molecular targetExperimental modelUltrasound imaging protocolKey findings
Leong-Poi et al.29Biotinylated lipid microbubbles bearing either mAb against αV-integrins or echistatin (using biotin–avidin linkage)Murine Matrigel plug modelHigh-power triggered power Doppler imagingFirst report of CEU for molecular imaging of angiogenesis. Direct observation of microbubble targeting in vivo. Showed peptide-based strategy comparable to mAb strategy for targeting
Leong-Poi et al.31Biotinylated lipid microbubbles bearing echistatin (using biotin–avidin linkage)Rodent model of acute hindlimb ischaemiaHigh-power triggered ultraharmonic imagingDemonstrated utility of CEU for molecular imaging of endogenous and growth factor-induced angiogenesis in ischaemic tissue
Stieger et al.30Biotinylated phospholipid microbubbles bearing echistatin (using biotin–avidin linkage)Rodent Matrigel plug modelLow-power real-time Cadence™ contrast pulse sequencing (CPS) imaging, with and without high-power destructive pulsesConfirmed utility of peptide-based strategy for microbubble targeting. Demonstrated the ability to use a low-power imaging technique (CPS)
Behm et al.32Lipid microbubbles with distearoylphosphatidylserine in the shell (targeting neutrophils) and biotinylated lipid microbubbles bearing either mAb against α5-integrins or VCAM-1 (using biotin–avidin linkage)Murine Matrigel plug model, and murine model of acute hindlimb ischaemiaHigh-power triggered power Doppler imaging in Matrigel plug model. High-power Cadence™ contrast pulse sequencing (CPS) imaging in hindlimb ischaemia modelDemonstrated utility of CEU molecular imaging to assess the inflammatory responses that contribute to vasculogenesis and arteriogenesis in response to ischaemia. Showed that multitarget imaging sessions was possible because of rapid clearance of freely circulating microbubbles from the blood pool, and the ability to destroy microbubbles

Microbubbles targeted against αV-integrins were first assessed by direct observation of their binding affinity within angiogenic vessels under physiological flow conditions in vivo, using intravital microscopy.29 By implanting fibroblast growth factor (FGF)-2-eluting pellets into the scrotum of mice, an angiogenic phenotype within the cremaster muscle microvasculature was created (Figure 1A). After intravenous injections of microbubbles, direct attachment of αV-targeted microbubbles (arrowheads) to the endothelial surface of arterioles and capillaries was observed (Figure 1B). Quantification of microbubble adherence performed 5 min after iv injection of 1 × 107 fluorescent microbubbles showed significantly greater adherence of αV-targeted when compared with control non-targeted microbubbles. Subsequently, CEU imaging was performed in an FGF-2-treated murine Matrigel plug model.29 Initially avascular, if supplemented by growth factors, angiogenesis is initiated and the Matrigel plug becomes vascularized from the surrounding tissues and blood vessels. Thus, it provides a physiologically relevant environment for studies of angiogenesis. Illustrated in Figure 2 are the images demonstrating marked contrast enhancement at the periphery of the Matrigel plug for microbubbles targeted by an mAb against αV-integrins and by the disintegrin echistatin, whereas contrast enhancement was low for control microbubbles bearing isotype control Ab.29 The results of this study were later confirmed by Stieger et al.30 using CEU and echistatin-bearing microbubbles to image angiogenesis in Matrigel plugs in rats. Importantly, these studies demonstrated that a targeting strategy using mAbs yielded similar results to a strategy using an RGD peptide, echistatin.

Figure 1

(A) Fluorescent microscopic images of the normal cremasteric microcirculation after intravenous injection of FITC–dextran to highlight the microvessels (left panel). After chronic exposure to FGF-2, there was a dense proliferation of microvessels, which demonstrated abnormal vascular permeability, as demonstrated by extravascular blush with FITC–dextran, characteristic of angiogenic vessels (right panel). Scale bar = 50 µm. (B) Intravital microscopy of the microcirculation of the cremaster muscle in FGF-2-treated mice demonstrating retention (arrowheads) of αV-targeted microbubbles in small arterioles (left panel) and in a capillary (right panel). Scale bar = 25 µm. (Reproduced with permission from Leong-Poi et al.29 Copyright Wolters Kluwer Health.)

Figure 2

Background-subtracted colour-coded CEU images reflecting signal from microbubbles retained in Matrigel plugs 15 min after intravenous injection of microbubbles targeted to angiogenesis. The colour-coding scale for background-subtracted acoustic intensity is shown at the bottom of each image. Although the signal from isotype control microbubbles was low, the signal from retained microbubbles bearing antibody vs. αVβ3 and echistatin was much greater, and localized to the periphery of plugs where angiogenesis had occurred. Graph below shows mean ± SD signal enhancement for retained microbubbles within the Matrigel plug. *P < 0.01 compared with MBC. (Reproduced with permission from Leong-Poi et al.29 Copyright Wolters Kluwer Health.)

This work was subsequently extended to image neovascularization in a clinically relevant rat model of chronic hindlimb ischaemia, using echistatin-bearing microbubbles.31 In this model, after unilateral iliac artery ligation blood flow to the ischaemic limb decreases immediately to ∼25% of normal resting values and over the next 2 weeks increases to ∼40% due to endogenous angiogenesis. After slow-release FGF-2 supplementation to the ischaemic hindlimb, neovascularization was significantly augmented, with a more rapid increase in perfusion over time. In this study, CEU molecular imaging of αVβ3 expression using an RGD peptide-conjugated microbubble agent was able to assess both the endogenous arteriogenic response to chronic occlusive PAD, as well as the therapeutic neovascularization response to prolonged growth factor administration with FGF-2. Figure 3 shows examples of targeted CEU images from ischaemic muscles immediately after (day 0) and 4 days after ligation. Signal enhancement from αVβ3-targeted microbubbles was greater and more diffuse in the FGF-2-treated muscle compared with the untreated ischaemic muscle where enhancement was located primarily around the region of the major vascular bundles, whereas signal was low for control microbubbles (data not shown).31

Figure 3

Background-subtracted colour-coded CEU images reflecting retention fraction of αVβ3-targeted microbubbles in ischaemic proximal hindlimb adductor muscles from control untreated and FGF-2-treated rats immediately after ligation (day 0) and at day 4 post-iliac artery ligation. Colour scales appear at bottom. In control untreated animals, signal from αVβ3-targeted microbubbles was seen to increase at day 4 post-ligation, in keeping with the endogenous angiogenesis. In comparison, in FGF-treated ischaemic muscle, there was an incremental increase in the signal from αVβ3-targeted microbubbles at day 4, consistent with a greater angiogenic response seen with FGF-2 treatment. (Reproduced with permission from Leong-Poi et al.31 Copyright Wolters Kluwer Health.)

Finally, Behm et al.32 used microbubbles targeted to activated neutrophils, α5-integrins, and vascular cell adhesion molecule (VCAM-1) to assess the inflammatory responses that contribute to vasculogenesis and arteriogenesis in response to ischaemia. In a murine Matrigel plug model, signal enhancement from α5-integrins and VCAM-1 coincided with early neovasculogenesis and correlated temporally with histology. In a murine hindlimb model of acute ischaemia, targeted CEU imaging demonstrated early signal enhancement for neutrophils, monocyte α5-integrins, and VCAM-1 at day 2 post-ligation at a time when tissue perfusion was still very low (Figure 4). Although targeted CEU signal from neutrophils declined at early time points, VCAM-1 and monocyte signal persisted to day 7 (Figure 4). In comparison, changes in tissue perfusion to the ischaemic hindlimb were more delayed, increasing slowly over time and peaking late at day 21 post-ligation.

Figure 4

CEU molecular imaging data for microbubbles targeted to activated neutrophils (A), α5-integrin (B), and VCAM-1 (C) at various time points after femoral artery ligation in mice. For each molecular target, mean ± SEM signal enhancement in the ischaemic limb is shown for targeted and control microbubbles. Data are normalized to muscle blood flow. Corresponding colour-coded examples of CEU molecular imaging with microbubbles directed to each molecular target at day 2 are shown below each graph. Colour scales appear at bottom. *P < 0.05 vs. control microbubbles. (Reproduced with permission from Behm et al.32 Copyright Wolters Kluwer Health.)

These studies illustrate several important aspects of CEU molecular imaging of angiogenesis in the setting of endogenous ischaemia and in assessing the response to strategies of therapeutic neovascularization. First, as microbubbles rely on vascular transit to reach their target, this technique may be relatively disadvantaged in ischaemic conditions where flow to the tissue of interest is reduced. In order to improve the sensitivity for the molecular imaging of angiogenesis in this setting, the signal from retained microbubbles was normalized to regional perfusion data from corresponding parametric perfusion images to obtain data on retention fraction. Similar adjustments to normalize the signal from targeted microbubbles for regional perfusion should be strongly considered for any CEU molecular imaging technique performed in the setting of reduced or low flow, such as chronic ischaemia. This would not necessarily be relevant in the setting of CEU molecular imaging of tumour angiogenesis, where regional blood flow within the tumour is increased. Secondly, CEU molecular imaging was able to detect integrin αVβ3 expression associated with vascular remodelling, both in response to ischaemia and to angiogenic growth factor administration, at time points prior to maximal changes in perfusion to the ischaemic muscle had occurred (Figure 5).31 Finally, CEU molecular imaging allowed the assessment of inflammatory processes that coincide with the appearance of functional microvessels, helping to differentiate the cellular components of the inflammatory response to ischaemia, and detecting expression of specific adhesion molecules on monocytes and endothelial cells that participate in vascular development and remodelling in response to acute ischaemia. These unique findings confirm that molecular imaging techniques can provide important information on underlying processes at the cellular and molecular level that precedes changes at a physiological or anatomic level. Thus, CEU molecular imaging may allow the early monitoring and accurate prediction of the angiogenic response to therapies, as well as provide mechanistic insights into the angiogenic response to ischaemia, highlighting the potential added value of molecular imaging data, over more traditional anatomic or functional data such as vessel density or perfusion.

Figure 5

(A) Normalized microvascular blood flow in the ischaemic muscle for untreated (filled circles) and FGF-2-treated (open circles) animals. Data from day 0 were obtained 1 h after arterial ligation. *P < 0.05 compared with data on day 0; P < 0.05 compared with corresponding data in untreated animals. (B) Mean background-subtracted acoustic intensity for αVβ3-targeted microbubbles in proximal hindlimb adductor muscles in ischaemic limb. Data are shown for untreated (filled circles) and FGF-2-treated (open circles) rats. *P < 0.05 compared with data on day 0; P < 0.05 compared with corresponding data for FGF-2-treated rats. The ultrasound signal from αVβ3-targeted microbubbles, both in untreated and FGF-2-treated muscle, is greatest at days 4 and 7 post-ligation, prior to maximal increases in muscle perfusion. (Reproduced with permission from Leong-Poi et al.31 Copyright Wolters Kluwer Health.)

Similar CEU molecular imaging techniques have been further validated in the evaluation of tumour angiogenesis, both for the early detection of tumours and for monitoring the response to treatment. Echistatin-bearing microbubbles targeted against αVβ3 have been used for the CEU molecular imaging of malignant gliomas in athymic rats.33 CEU signal from retained αVβ3-targeted microbubbles was localized mainly to the periphery of gliomas, where tumour vasculature and vascular expression of αV-integrin were most prominent. Several other molecular targets for the CEU imaging of angiogenesis have been tested in tumour models, including (i) microbubbles targeted against tumour vasculature by the attachment of a unique tumour endothelial cell-specific binding peptide (arginine–arginine–leucine) for imaging of prostate tumors,34 (ii) microbubbles bearing an mAb against vascular endothelial growth factor receptor-2 (VEGFR2) for imaging breast cancer35 and human melanomas,36 and (iii) dual-targeted microbubbles against both αVβ3 and VEGFR2 for the molecular imaging of angiogenesis in a human ovarian cancer xenograft tumour model in mice.37 Studies have also demonstrated that CEU molecular imaging of angiogenesis can track the regression of tumour vasculature in response to antitumour therapies. Using microbubbles targeted to αVβ3 and VEGFR2, Palmowski et al.38 followed the regression of tumour vasculature in response to matrix metalloproteinase inhibitors in squamous cell carcinomas in athymic mice, whereas Korpanty et al.39 demonstrated that targeted CEU and microbubbles targeted to endoglin, VEGF–VEGFR complex, and VEGFR2 could monitor the vascular effects of anti-VEGF and gemcitabine treatment on pancreatic adenocarcinoma in mice. Thus, similar to CEU molecular imaging of therapeutic angiogenesis, microbubbles targeted to tumour angiogenic vessels can also be used to longitudinally assess responses to anti-angiogenic cancer therapies at the molecular and cellular level.

5. CEU molecular imaging of cell therapy

With the rapid progression of research into stem or progenitor cell therapy to treat cardiovascular diseases, there is a need to develop imaging modalities to track progenitor cells in vivo. Although many other imaging techniques have been evaluated for tracking the fate of stem cells, the ability to track delivered cells using CEU has not been previously investigated. Optical imaging methods, such as bioluminescence40,41 and fluorescence imaging,42 have been utilized for stem cell tracking. Bioluminescence relies on the stable transfection of cells with genes such as luciferase to generate light that can then be detected in vivo, whereas fluorescence utilizes organic or organic/inorganic hybrids as cellular contrast agents. Both techniques have limited clinical potential due to poor tissue penetration and low depth of detection, making them more suitable for small animal studies and ex vivo molecular imaging. Radionuclide (SPECT and PET) techniques track cells either by direct loading of a radiometal, enzymatic conversion and retention of a radioactive substrate or receptor-mediated binding,27,43 whereas MRI techniques target cells by loading them prior to delivery with gadolinium-containing contrast agents or superparamagnetic iron oxide nanoparticles.44,45 Radionuclide techniques have the advantage of high sensitivity, whereas MRI techniques hold the edge in superior spatial resolution.

Contrast ultrasound techniques have now been developed that would potentially allow the tracking of delivered progenitor cells in vivo. Similar to other techniques to track stem cells, due to the lack of reliable and specific stem cell biomarkers that can be tracked with current imaging tools, cells have to be manipulated ex vivo prior to injection to allow their in vivo tracking. For CEU molecular imaging of stem cells, two potential strategies exist (Figure 5). The first strategy relies on the generation of a unique marker on the surface of stem cells, which can then be targeted by attaching the specific ligand to this target onto the microbubble surface. The second method involves the ‘loading’ of cells with intact microbubbles prior to delivery (Figure 5). The CEU method to image these targeted cells will differ depending on the labelling strategy. For the cell surface marker technique, CEU molecular imaging will be similar to that for the targeted imaging of angiogenesis, whereby after cell delivery and engraftment, targeted microbubbles would be injected and allowed to circulate and bind to their target receptors on engrafted cells (Figure 5). Ultrasound imaging could then be performed to detect the signal from adhered microbubbles on the surface of delivered and engrafted cells. As microbubbles remain purely intravascular after intravenous administration, this technique would only allow the targeted imaging of cells engrafted ‘within’ the vessel walls. However, as long as the specific marker remains expressed on the cell surface, repeat imaging studies could be performed, thus allow serial assessments and longitudinal tracking of cell engraftment. For the second technique, after cell delivery CEU imaging could be performed without additional administration of microbubbles. Ultrasound imaging would then detect the signal generated from microbubbles that are present within engrafted cells. Similar to SPECT and MRI techniques, this method would have the theoretical advantage of tracking cells, regardless of their location, intravascular, or extravascular. Depending on the ultrasound imaging parameters used, microbubbles within cells may be disrupted during their imaging. This would then negate the possibility of sequential studies to track cell engraftment over time. In addition, prior to clinical application, the effects of intracellular microbubble disruption on cell viability would have to be defined.

To date, two preliminary studies of targeted CEU for imaging cell therapy have been performed. In this Review Focus on Molecular Imaging in the Cardiovascular System in Cardiovascular Research, Kuliszewski et al.46 investigated the CEU molecular imaging of EPC engraftment in vivo, using the first of the two techniques discussed above. Cultured bone marrow-derived EPCs were transfected to express a unique marker protein, H-2Kk, on the cell surface. This strategy of ex vivo cell transfection to label stem cells with an imaging biomarker has been used by other techniques, including bioluminescence imaging where cells are transfected with a reporter gene, such as Firefly luciferase.40 By attaching the mAb against H-2Kk onto the outer surface of microbubbles, EPC-targeted microbubbles were created. The binding characteristics were tested using an in vitro flow chamber system similar to that used in the development of other targeted microbubbles, showing strong and selected binding to plated transfected EPCs, with minimal binding to control mock-transfected EPCs. In vivo CEU molecular imaging of EPCs engrafted into the vasculature within Matrigel plugs was then performed. Using anti-H-2Kk microbubbles, targeted CEU enhancement was noted within the vascularized portion of plugs supplemented with H-2Kk-transfected EPCs with very little within control plugs supplemented with mock-transfected EPCs, whereas control microbubbles yielded low signal from both H-2Kk- and mock-transfected EPC-supplemented plugs (Figure 6).

Figure 6

Schematic diagram of potential strategies for CEU-targeted imaging of stem/progenitor cells, such as EPCs. One strategy involves manipulation/transfection of EPCs to express a specific marker protein on their surface. Microbubbles (MB) targeted to engrafted EPCs could be constructed by the attachment of the ligand/antibody targeted against the marker protein on the microbubble surface (left aspect of diagram), which when administered intravenously could bind to EPCs that are engrafted within the vasculature, and be imaged by CEU. The other strategy involves manipulating EPCs to fully engulf MB, prior to delivery. Once delivered and engrafted, ultrasonic imaging could then detect MBs present within engrafted EPCs (right aspect of diagram).

The second CEU molecular imaging technique to track stem cells has been studied by Cui et al.47 In their study, MSCs were incubated with perfluorocarbon gas-filled microbubbles, resulting in their internalization within the cells. Upon CEU imaging, the signal from microbubble-filled MSCs was significantly greater than unlabelled MSCs, confirming the ability to detect the microbubble labelled MSCs with ultrasonic imaging. These studies are the first to describe CEU molecular imaging of progenitor cells and demonstrate the potential of this technique for tracking engraftment of progenitor cells after their delivery (Figure 6).

6. Conclusions and future directions

As presented in this review, studies to date have validated the use of CEU molecular imaging techniques for the evaluation of angiogenesis, and these principles have now been extended to image stem cells in preliminary experiments that at the very least provide proof of concept. However, when compared with progress made by other imaging techniques for the evaluation of angiogenesis and cell therapy, such as radionuclide techniques—SPECT and PET for the molecular imaging of angiogenesis, or optical imaging and MRI for stem cell tracking, the development of CEU molecular techniques for the evaluation of angiogenesis and cell therapies is lagging behind. For angiogenesis imaging, the majority of work has been performed in tumour models of angiogenesis, with only two studies performed in a model of vascular insufficiency.31,32 Furthermore, there are currently no studies of CEU molecular imaging of angiogenesis in cardiac models of ischaemia or infarction. Although recent preliminary experiments provide data on CEU molecular imaging to track cell therapy, future studies still need to be performed to further validate this technique, in particular in models of ischaemia and infarction.

The development of CEU techniques for the molecular imaging of therapeutic angiogenesis and cell therapies in the setting of ischaemia and infarction faces unique challenges, which are also inherent to other imaging modalities. These challenges are less relevant in the setting of imaging tumour angiogenesis. In the ischaemic setting, the magnitude of response to strategies of therapeutic angiogenesis, both gene- and cell-based therapies, has been relatively modest. Certainly, the negative clinical trials of gene therapy for angiogenesis to date clearly illustrate this. Similarly, the magnitude of early cell engraftment after delivery appears to be fairly modest and decreases significantly over time. Studies have demonstrated that the benefit of cell therapy may not solely be attributable to the physical engraftment of cells into neovessels, but may be explained by the concomitant release of pro-angiogenic factors and cytokines, through paracrine mechanisms. Although ongoing research into optimizing gene delivery techniques for angiogenesis and increasing cell engraftment continues, current molecular imaging techniques need to be highly sensitive to detect these relatively rare events.

For CEU, additional improvements in microbubble and contrast-specific ultrasound imaging technology have the potential to provide the high sensitivity required to image pro-angiogenic responses and track cell therapy. Although commercially available microbubble contrast agents are approved for left ventricular or blood pool opacification, no agents are currently approved for myocardial perfusion. Microbubble contrast agents for molecular imaging have been developed for pre-clinical use; however, Phase I human clinical trials have yet to be performed. With optimization of imaging algorithms that suppress tissue signal, the sensitivity of ultrasound for the detection of microbubble agents is high, with reports of ‘single’ bubble imaging.19 In addition, studies suggesting that signal subtraction and decorrelation detection can allow the differentiation of adherent αVβ3-targeted microbubbles from free-flowing microbubbles may lead to further increases in the sensitivity for the detection of retained targeted microbubbles.48

In addition to improving the sensitivity for the detection of targeted microbubbles, newer higher frequency imaging algorithms and the development of novel microbubble agents designed for high transmit frequency imaging will be important in optimizing the spatial resolution of CEU molecular imaging. The majority of studies of CEU molecular imaging of angiogenesis have employed high-power destructive techniques, in part to maximize the signal from the relatively small proportion of retained targeted microbubbles. Recent studies have validated the use of low-power non-destructive imaging sequences, such as cadence contrast pulse sequencing,30 for CEU molecular imaging of angiogenesis. Further refinements of these real-time low-power ultrasound imaging techniques to detect targeted microbubbles will help facilitate translation of molecular imaging of myocardial angiogenesis in patients.

If such concepts, to increase sensitivity, optimize spatial resolution, and perform CEU molecular imaging in real time, can be extended to targeted echocardiographic imaging for myocardial angiogenesis and cell tracking in response to therapeutic interventions, the promise of CEU molecular imaging for the evaluation of angiogenesis and cell therapy for cardiovascular diseases may yet be fulfilled, whereby the response to novel gene- and cell-based regenerative therapies in patients can be evaluated, and treatment can be tailored based upon CEU molecular imaging data.

Conflict of interest: none declared.


This work was supported by an Operating Grant from the Canadian Institutes of Health Research (MOP 89836), Ottawa, Ontario, Canada. H.L.-P. is supported by a Phase II Clinician Scientist Award from the Heart and Stroke Foundation of Ontario, Canada and an Early Researcher Award from the Ministry of Research and Innovation, Ontario, Canada.


View Abstract